What it shows:
A metal under stress will not fracture straight away, but will deform plastically due to the dislocation of crystal boundaries; this is called creep.
How it works:
Here we use lead as the test sample because there is significant creep compared to other metals. The lead is loaded (see fig.1) to a value that is just below the breaking stress of the sample. When creep occurs, the lead is drawn thinner at its weakest point (called 'necking', see fig.2) until its reduced cross-sectional area causes the sample to exceed its breaking stress and fracture.
Setting it up:
The lead sample is milled down to shape from a 0.3 × 2 × 16cm plate. To create the curved collar, drill out with a 1cm drill. For ease of machining (lead isn’t easy to machine) clamp three samples together; this also means you get similar pieces for subsequent demos. The lead is held at either end with parallel jawed (Kant Twist®) clamps, which give even gripping of the sample and allow a load to be suspended on a center line so there is no torque on the sample. The top clamp can be fixed to a bench using a right angle layout tool with G-clamp. When tightening the clamps on the lead, be careful not to twist its waist. For the load, use the slotted weights; 19kg ± 1kg will give a test lifetime of 10-20 minutes (see Comments). To get 20kg on the weight holder you may need to duck-tape a couple of 2kgs to the sides of their comrades. Front lighting will help visibility, and maybe a camera to zoom in.
Very inconsistent results makes this a hit-or-miss demo. Seemingly identical samples tested one after another show one failing at 30 seconds and the other hanging around all afternoon. The action is also unclear to the audience.